The Plant Journal (2015) 83, 719–731

doi: 10.1111/tpj.12922

The deubiquitinating enzyme AMSH1 is required for rhizobial infection and nodule organogenesis in Lotus japonicus
ski1, Euan K. James2, Niels Sandal1, Erika Isono3, Jens Stougaard1 and Anna Małolepszy1, Dorian Fabian Urban 1, Stig Uggerhøj Andersen * 1 Centre for Carbohydrate Recognition and Signalling, Department of Molecular Biology and Genetics, Aarhus University, Gustav Wieds Vej 10, DK–8000, Aarhus C, Denmark, 2 The James Hutton Institute, Invergowrie Dundee DD2 5DA, UK, and 3 Department of Plant Systems Biology, Technische Universita€ t Mu€ nchen, D–85354 Freising, Germany Received 22 May 2015; accepted 22 June 2015; published online 28 june 2015. *For correspondence (e-mail [email protected]). Accession numbers: The Amsh1 sequences have been submitted to Genbank under accession number KR270441.

SUMMARY Legume–rhizobium symbiosis contributes large quantities of fixed nitrogen to both agricultural and natural ecosystems. This global impact and the selective interaction between rhizobia and legumes culminating in development of functional root nodules have prompted detailed studies of the underlying mechanisms. We performed a screen for aberrant nodulation phenotypes using the Lotus japonicus LORE1 insertion mutant collection. Here, we describe the identification of amsh1 mutants that only develop small nodule primordia and display stunted shoot growth, and show that the aberrant nodulation phenotype caused by LORE1 insertions in the Amsh1 gene may be separated from the shoot phenotype. In amsh1 mutants, rhizobia initially became entrapped in infection threads with thickened cells walls. Some rhizobia were released into plant cells much later than observed for the wild-type; however, no typical symbiosome structures were formed. Furthermore, cytokinin treatment only very weakly induced nodule organogenesis in amsh1 mutants, suggesting that AMSH1 function is required downstream of cytokinin signaling. Biochemical analysis showed that AMSH1 is an active deubiquitinating enzyme, and that AMSH1 specifically cleaves K63-linked ubiquitin chains. Post-translational ubiquitination and deubiquitination processes involving the AMSH1 deubiquitinating enzyme are thus involved in both infection and organogenesis in Lotus japonicus. Keywords: Lotus japonicus, nodulation mutant, AMSH1, deubiquitination, symbiosis, LORE1.

INTRODUCTION Many legumes are rich sources of protein due to their symbiosis with nitrogen-fixing rhizobia accommodated in root nodules. The symbiotic signaling process is initiated when rhizobia secrete nodulation (Nod) factors upon sensing (iso) flavonoids produced by compatible legumes. The Nod factors consist of N–acetyl-D–glucosamine oligomers linked by b–1,4 bonds, in which the N–acetyl group on the non-reducing end is replaced by a fatty acid (Lerouge et al., 1990; Spaink et al., 1991). Genetic studies of loss-of-function and gain-of-function mutants in Lotus japonicus (Lotus) and Medicago truncatula (Medicago) show that the Lotus Nod factor receptors NFR1 and NFR5 and the corresponding proteins LYK3 and NFP in Medicago (Limpens et al., 2003; Madsen et al., 2003; Radutoiu et al., 2003; Arrighi et al., 2006; Mulder et al., 2006; Smit et al., 2007) are pivotal for perception of rhizobial Nod factors, and NFR1 and NFR5 bind Nod © 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd

factors from compatible bacteria in vitro with high affinity (Broghammer et al., 2012). In the susceptible root zone, Nod factor perception then leads to initiation of a nodulation signaling cascade, which bifurcates into branches promoting epidermal root hair infection and cortical nodule organogenesis (Kouchi et al., 2010; Madsen et al., 2010). Promotion of epidermal infection relies on the activities of the pectate lyase Nodulation pectate lyase (NPL), the ubiquitin ligase CERBERUS, the Nck-associated protein 1 (NAP1), 12IF-specific p53 inducible RNA (PIR1) and Actin-related protein component 1 (ARPC1) proteins required for actin rearrangement, and the transcription factors Nodule inception protein (NIN), Nodulation signaling pathway 1 (NSP1) and Nodulation signaling pathway 2 (NSP2) (Schauser et al., 1999; Kalo et al., 2005; Smit et al., 2005; Heckmann et al., 2006; Murakami et al., 2006; Marsh et al., 2007; Yano et al., 2008, 2009; Yokota et al., 2009; Hossain et al., 2012; Xie et al., 2012). 719

720 Anna Małolepszy et al. An overlapping set of proteins act downstream of NFR1 and NFR5 to trigger the organogenesis process in the root cortex. The leucine-rich repeat receptor kinase Symbiosis receptor-like kinase (SymRK), and three nucleoporins (NUP133, NUP85 and NENA), together with CASTOR and POLLUX cation channel proteins (Schauser et al., 1998; Szczyglowski et al., 1998; Endre et al., 2002; Kawaguchi et al., 2002; Stracke et al., 2002; Imaizumi-Anraku et al., 2005; Kanamori et al., 2006; Saito et al., 2007), are essential for the nuclear calcium-spiking signal that is in turn interpreted by Ca2+ and calmodulin-dependent receptor kinase (CCaMK) (Levy et al., 2004; Mitra et al., 2004; Tirichine et al., 2006a). Subsequently, CCaMK activates organogenesis via cytokinin signaling through the LHK1 receptor (Tirichine et al., 2006b, 2007; Murray et al., 2007), but also cross-signals with the infection pathway by phosphorylating and activating the transcription factor CYCLOPS (Yano et al., 2008; Liao et al., 2012; Singh et al., 2014). Further downstream, the transcription factors NIN, NSP1 and NSP2 (Schauser et al., 1999; Kalo et al., 2005; Smit et al., 2005; Heckmann et al., 2006; Murakami et al., 2006; Marsh et al., 2007) relay the organogenesis signal (Madsen et al., 2010). Several of the key nodulation proteins mentioned above are associated with ubiquitination, either because they are putative targets for ubiquitination, or because they share sequence similarity with ubiquitin ligases. CERBERUS belongs to the latter category, whereas the Medicago Nod factor receptor LYK3 phosphorylates the E3 ubiquitin ligase Plant U–box protein 1 (PUB1) (Mbengue et al., 2010), and SymRK has been suggested to interact with the E3 ubiquitin ligases SymRK INTERACTING E3 UBIQUITIN LIGASE (SIE3) in Lotus, and SEVEN IN ABSENTIA (SINA) in Medicago (Den Herder et al., 2008, 2012; Herve
et al., 2011; Yuan et al., 2012). Ubiquitination is a reversible process that regulates protein degradation, trafficking, DNA repair, apoptosis and signal transduction, making it an important posttranslational regulatory mechanism. It involves linking the C–terminal glycine residue of ubiquitin to lysine residues of specific protein targets, in many cases followed by ubiquitin polymerization (Kimura and Tanaka, 2010). The fate of the target protein is determined by the identity of the lysine residue involved in poly-ubiquitin chain formation. If the poly-ubiquitin chain is generated by K48 linkage, the substrate is targeted to the proteasome for degradation. On the other hand, formation of K63-linked chains promotes DNA repair, endocytosis and vesicular trafficking, or ribosomal protein synthesis (Mukhopadhyay and Riezman, 2007; Sato et al., 2008; Ye and Rape, 2009). Accurate control of protein ubiquitination status is crucial for the cell, which is why deubiquitinating enzymes (DUBs) play an important role in cellular homeostasis (Kimura and Tanaka, 2010). DUBs are isopeptidases that detach ubiquitin from target proteins. They are divided into

five groups: ubiquitin-specific proteases, ubiquitin C–terminal hydrolases, Otubain proteases, Machado–Joseph disease proteases and JAB1/MPN/Mov34 metalloenzymes (JAMMs) (Isono and Nagel, 2014). All DUBs belong to the cysteine protease group, except for the JAMMs, which are zinc metalloproteases (Amerik and Hochstrasser, 2004; Nijman et al., 2005; Sato et al., 2008; Isono and Nagel, 2014). The metalloproteases contain an MPN+ domain with a conserved JAMM motif EXnHXHX10D, which is believed to be the active site involved in metal binding (Maytal-Kivity et al., 2002; Tran et al., 2003). In Arabidopsis, three homologs of ASSOCIATED MOLECULE WITH THE SH3 DOMAIN OF STAM (AMSH), which belong to the JAMM zinc metalloprotease class, have been identified: AMSH1, AMSH2 and AMSH3. Their catalytic MPN+ domain is homologous to the MPN+ domain in human AMSH. AMSH1 and AMSH3 interact with ENDOSOMAL SORTING COMPLEX REQUIRED FOR TRANSPORT III (ESCRT–III), which is involved in sorting of ubiquitinated membrane proteins to multivesicular bodies (Isono et al., 2010; Katsiarimpa et al., 2011, 2013). Recently, it was shown that AMSH3 is involved in the degradation of AvrPtoB-activated CHITIN ELICITOR RECEPTOR KINASE 1 (CERK1) (Katsiarimpa et al., 2014). Despite the aforementioned associations between proteins from the nodulation pathways and the ubiquitination machinery, the role of ubiquitination in nodulation signaling is not well understood. Using an unbiased forward-genetics approach, we screened the LORE1 collection (Fukai et al., 2012; Urbanski et al., 2012) for mutants with aberrant nodulation phenotypes, and found that loss of a gene encoding the deubiquitinating enzyme AMSH1 prevented nodulation. Our analysis showed that Lotus AMSH1 is an active DUB with specificity for K63-linked ubiquitin, and highlights the importance of post-translational protein modification and ubiquitin signaling in the developmental events leading to root infection and nodule organogenesis. RESULTS Identification of two independent LORE1 mutants with similar aberrant nodulation phenotypes Following the successful establishment and annotation of a small-scale LORE1 insertion mutant population (Fukai et al., 2012; Urbanski et al., 2012), we performed a forward-genetic screen for nodulation-deficient mutants. We identified a LORE1 family segregating plants without mature nodules, although no insertions in known nodulation genes were registered in the LORE1 database. In addition to displaying arrest of nodule development at the primordium stage, the mutant plants showed severely stunted shoot growth, but root growth was less affected (Figure 1a–e). Additionally, the mutants did not survive until seed set. Initially, we were unable to identify the causative insertion based on our LORE1 sequencing data, as none of the

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

AMSH1 promotes infection and organogenesis 721

(a)

of the Amsh1 gene in an independent LORE1 mutant line (amsh1–1) (Figure 1f). In addition to Amsh1, the Lotus version 2.5 genome contains two additional Amsh homologs that appear to be putative orthologs of the Arabidopsis AMSH2 and AMSH3 genes (Figure S1).

(b)

LORE1 insertions in the Amsh1 gene cause the aberrant nodulation phenotype

(c)

(d)

(e)

(f)

Figure 1. amsh1 nodulation phenotype. (a,b) Plants grown on quarter-strength B&D plates, 6 weeks after inoculation with M. loti rhizobia. (c–e) Plant root systems at 6 wpi. Fully developed nodules are indicated by arrows, and nodule primordia are indicated by arrowheads. Scale bar = 0.5 mm. (f) Structure of the L. japonicus Amsh1 gene (GenBank accession number KR270441). The exon–intron structure was obtained by alignment of the sequenced full-length Amsh1 cDNA with Lotus contigs assembled from short Illumina reads (data not shown). The Amsh1 gene has 14 exons, covers approximately 6.2 kb of genomic sequence, and encodes a protein of 58 kDa. LORE1 insertion sites are shown for amsh1 mutant alleles.

annotated insertions co-segregated with the phenotype. After performing sequence-specific amplification polymorphism analysis optimized for LORE1 insertion detection (Urbanski et al., 2013), followed by Sanger sequencing of candidate PCR products, we detected a co-segregating insertion (amsh1–2) in exon 8 of the Amsh1 gene. Because Amsh1 was not included in the Lotus version 2.5 reference genome sequence, we then cloned and sequenced the fulllength Amsh1 cDNA sequence, and identified the Amsh1 genomic DNA sequence by aligning the cDNA sequence with a set of Lotus contigs assembled from short Illumina reads (data not shown). After re-running the analysis of the LORE1 sequencing data, including the Amsh1 genomic sequence, we discovered an additional insertion in exon 1

The phenotypes of the two independent amsh1 alleles were compared. As the homozygous amsh1 mutants did not produce seeds, offspring from heterozygous plants were used in all analyses. We found that all plants with stunted shoots and small nodule primordia carried homozygous LORE1 insertions in the Amsh1 gene. To confirm that the two LORE1 insertions in the Amsh1 gene were the causal mutations, heterozygous amsh1–1 and amsh1–2 plants were crossed to test for non-complementation in the F1 generation. amsh1–1 amsh1–2 heteroallelic individuals displayed phenotypes indistinguishable from the homozygous amsh1–1 and amsh1–2 mutants (Figure S2). amsh1–1 and amsh1–2 are therefore allelic, and we conclude that the LORE1 insertions in the Amsh1 gene are the causal mutations. To further validate correct identification of the causal mutations, we identified six independent LORE1 lines following expansion of the LORE1 population. These additional alleles (amsh1–3 to amsh1–8; Figure 1f) all displayed phenotypes very similar to the amsh1–1 and amsh1–2 alleles, re-confirming correct identification of the causal mutations (Figure S3). The amsh1 mutants accumulate ubiquitin conjugates in vivo Alignment of the conceptual Lotus AMSH1 amino acid sequence with AMSH1 protein sequences from human and Arabidopsis AMSH metalloproteases showed that they share 42% and 64% overall amino acid sequence identity, respectively (McCullough et al., 2004; Isono et al., 2010), suggesting that the Lotus AMSH1 protein may be a metalloprotease enzyme with DUB activity (Figure S4). In contrast to ubiquitin ligases, DUBs cleave ubiquitin chains from target proteins, and loss of DUB function may lead to accumulation of ubiquitin conjugates (Wilkinson et al., 1995; Amerik et al., 1997). To test whether amsh1 mutants accumulated ubiquitinated proteins, we used an anti-ubiquitin antibody to detect ubiquitin conjugates in total extracts from wildtype, amsh1–2, heterozygous Amsh1/amsh1, nfr1–1 and nfr5–2 seedlings. Cyclin-dependent kinase (CDC2 kinase) (Riabowol et al., 1989; Madrid et al., 2007) was used as a loading control. The amsh1 mutant clearly had the highest level of ubiquitin conjugates (Figure 2a), and there were no apparent differences between protein samples from the remaining genotypes.

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

722 Anna Małolepszy et al. clude that Lotus AMSH1 is an active DUB that specifically cleaves K63-linked poly-ubiquitin chains in vitro. These results are consistent with the previously reported K63 chain specificities of human AMSH and Arabidopsis AMSH1 (McCullough et al., 2004; Katsiarimpa et al., 2013).

(a)

AMSH1 acts in Lotus roots to promote nodulation

(b)

Figure 2. AMSH1 deubiquitination activity. (a) Western blot using an anti-ubiquitin antibody (P4D1) to detect ubiquitin conjugates in protein extracts from plant seedlings. An anti-CDC2 kinase antibody was used to detect the CDC2 kinase loading control. wt, Gifu; het L1, heterozygous Amsh1/amsh1 plant derived from the amsh1–2 line; wt L1, wild-type plant derived from the amsh1–2 line. (b) Deubiquitination assay using linear, K48- and K63-linked ubiquitin chains as substrates. The ubiquitin chains were incubated with or without Lotus AMSH1 for 0 or 120 min. The numbers 1–5 indicate the number of ubiquitin units in the chains. An anti-ubiquitin antibody (P4D1) was used to detect ubiquitin chains, and an anti-GST antibody was used to detect the presence of the GST–AMSH1 protein.

Lotus AMSH1 specifically cleaves K63-linked ubiquitin chains Ubiquitin molecules may be linked through various lysine (K) residues to form poly-ubiquitin chains, and signaling responses may be modulated by variations in the patterns of lysine links (Komander and Rape, 2012). DUBs act by cleaving the amide bond between the lysine and the C–terminal glycine of a linked ubiquitin (Amerik and Hochstrasser, 2004). To examine whether AMSH1 is an active DUB, the fusion protein GST–LjAMSH1 was expressed in the Escherichia coli Rosetta strain, purified and tested for DUB activity. K48-linked and linear ubiquitin chains remained intact upon LjAMSH1 treatment, whereas K63-linked chains were cleaved to ubiquitin monomers (Figure 2b). We con-

As the amsh1 mutants displayed a severely stunted shoot phenotype, and as Amsh1 was expressed strongly across all Lotus tissues (Figure S5), the amsh1 mutation may indirectly affect root growth and progression of nodulation through its effect on shoot function and/or delivery of photosynthates to the root. To determine whether this was the case, we first examined root and shoot growth rates in the wild-type and in the amsh1–1 and amsh1–2 mutants. The growth rate was measured at 1, 2 and 3 weeks for plategrown plants with or without rhizobial inoculation (Figure 3 and Figure S6). Although nitrogen starvation inhibited shoot growth in uninoculated wild-type plants, the amsh1 shoot growth rate was significantly lower than that of the wild-type, both with and without inoculation (Figure 3a,c). In contrast, root growth rates were similar across all genotypes independent of inoculation, despite the short initial lengths of the amsh1 roots (Figure 3b,d). With respect to lateral root formation, there were no significant differences between amsh1 mutants and the wild-type with respect to the number of lateral roots produced per cm of main root, but the total number of lateral roots per plant was reduced (Figure 3e,f). Despite the relatively severe amsh1 shoot phenotype, root growth and development were thus not strongly affected, suggesting that the indirect shoot effect on the root phenotype is minimal in plate-grown amsh1 mutants. As further testimony to the relatively normal general functionality of amsh1 roots, they supported mycorrhizal colonization (Figure S7), which distinguishes amsh1 from the Medicago vapyrin mutant (Murray et al., 2011). Although root growth was not severely affected, Nod factor signaling and/or nodule development may potentially still be affected by limited shoot growth and functionality. To examine this possibility, we used Agrobacterium rhizogenes-mediated transformation to generate transgenic roots expressing the wild-type Amsh1 cDNA sequence. Using this system, the transgene is exclusively expressed in the transformed roots, allowing us to separate shoot and root effects. Expressing Amsh1 from the CaMV 35S promoter restored development of large pink nodules in 17 of 26 amsh1 mutants, although their shoots remained stunted compared to the wild-type controls (Figure 4, Table S1 and Figure S8). Vibratome sections (Figure 4) showed that these large pink nodules were fully infected and appeared indistinguishable from wild-type nodules at the same developmental stage. However, closer inspection of the complemented nodules using electron microscopy showed that, although symbiosomes were detected in the complemented amsh1 nodules,

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

AMSH1 promotes infection and organogenesis 723

(a)

(b)

pink nodules had significantly longer shoots than amsh1 mutants without pink nodules (P = 0.0048, Student’s t test) (Figure 4e), indicating that the pink nodules on the complemented plants had contributed fixed nitrogen to the amsh1 seedling shoots. Despite the premature symbiosome senescence observed, transformation of roots with the 35S: Amsh1 construct thus restored both infection and, at least partly, nitrogen fixation in the amsh1 mutants, and we conclude that AMSH1 acts in the root to promote nodulation. Infection thread progression is delayed in amsh1 mutants

(c)

(d)

(e)

(f)

Figure 3. amsh1 shoot and root growth rates. (a) Shoot lengths of plate-grown seedlings. (b) Root lengths of plate-grown seedlings. (c) Mean shoot growth rates between week 1 and week 3. (d) Mean root growth rates between week 1 and week 3. (e) Mean number of lateral roots per plant at 3 weeks. (f) Mean number of lateral roots per cm of root at 3 weeks. Asterisks indicate statistically significant differences compared with the wild-type (*P < 0.05, **P < 0.01, ***P < 0.001, Student’s t test). Error bars indicate SEM. Twenty Gifu plants, 17 amsh1–1 plants and 15 amsh1–2 plants were assayed.

they were less frequent than in wild-type nodules, and appeared to be partly degraded (Figure S8). To determine whether the degraded symbiosomes observed at 6 weeks post-inoculation (wpi) in the complemented plants had been functional at earlier stages, we quantified the shoot lengths of amsh1 plants in the complementation experiment. Complemented amsh1 plants with

To determine the exact nature of the amsh1 nodulation defect, we characterized the formation of infection threads (ITs) in root hairs. These ITs represent the early stages of rhizobial infection that allow the bacteria to traverse plant epidermal cells. The earliest morphological response to rhizobia is root hair curling, which we observed in both the wild-type and amsh1 mutants (Figure S9a–c). We then used DsRed-labeled Mesorhizobium loti to track IT progression, and categorized ITs into three groups: (1) incipient (microcolonies/short ITs not progressing), (2) elongating (incomplete traversal of root hair cells) and (3) long (complete traversal of root hair cells) (Figure 5a–c). At 8 days post-inoculation (dpi), the wild-type had approximately 30 long ITs per cm (Figure 5d), and also showed a number of incipient and elongating ITs. In contrast, the number of long ITs was significantly reduced to approximately one IT per cm in amsh1–1 and amsh1–2 (Figure 5d). Instead, we found mainly incipient or elongating ITs in amsh1–1 and amsh1–2, some with aberrant morphology (Figure S9d–g). At 21 dpi, approximately 30 and 10 long ITs per cm were observed in the wild-type and in the amsh1–1 and amsh1–2 mutants, respectively. The number of incipient and elongating ITs was similar in amsh1–1, amsh1–2 and wild-type plants, suggesting that the infection process was progressing at a reduced speed in the amsh1 mutants (Figure 5d). It thus appears that amsh1–1 and amsh1–2 are able to perceive rhizobial signals and initiate the infection process, but that IT progression is aberrant and delayed, resulting in decreased infection. amsh1 infection threads display aberrant morphology and bacterial release The slower infection process in amsh1 mutants was also reflected at the level of nodule organogenesis (Figure 5e). Wild-type plants produced nodule primordia within days of infection, and functional pink nitrogen-fixing nodules were present at 11 dpi. In contrast, none of the amsh1 mutants had fully developed mature nodules at 6 wpi (Figures 5e and 6). To investigate whether the infection status of the amsh1 primordia explained the lack of mature nodules, we visualized the infection of wild-type nodules and amsh1 nodule primordia 6 weeks after inoculation with M. loti

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

724 Anna Małolepszy et al. Figure 4. 35S::Amsh1 complementation of amsh1–1 in transgenic roots. (a) Plants with transgenic roots at 6 wpi. Arrowheads indicate nodule primordia; arrows indicate pink well-developed nodules. Scale bars = 1 cm. (b) Close-up views of root and nodules/nodule primordia at 6 wpi. The top panels are light micrographs. The red signal in the middle panels indicates DsRed-labeled M. loti rhizobia, and the lower panels are overlays of the two panels above. Scale bars = 0.5 mm. (c) Vibratome sections of wild-type nodules and amsh1 nodule primordia at 6 wpi. The purple signal indicates DsRed-labeled rhizobia. The blue signal represents cell-wall autofluorescence. The dashed white lines indicate regions with branched infection threads typical of amsh1, while dashed yellow lines indicate groups of normally infected cells. Infected cells were rarely seen in amsh1–1. Scale bars = 50 lm. (d) Close-up views of the regions indicated by white boxes in (c). Scale bars = 100 lm. The constructs used for transformation and the plant genotype are indicated at the top and bottom of the panels, respectively. (e) Shoot lengths of amsh1 plants. The mean shoot length of 35S::Amsh1 complemented amsh1 plants with pink nodules was significantly larger than the mean shoot length of 35S::Amsh1-transformed amsh1 plants without pink nodules and amsh1 plants transformed with an empty vector control (P = 0.0048, Student’s t test).

(a)

(b)

c). In some amsh1–1 and amsh1–2 primordia, ITs had entered cortical cells, and these ITs were irregular, bulbous and enclosed in unusually thick cell wall-like structures (Figure 6b,c). Release of bacteria into the plant cells was not observed in the amsh1 mutants. At 6 wpi, the infected cells of wild-type nodules showed symbiosomes containing up to two or three bacteroids, while in the occasional larger amsh1–1 and amsh1–2 nodule primordia, a higher number of branching ITs with aberrant morphology were present (Figures 4c and 6d,e). Release of bacteria was observed for a few of the older amsh1 mutant primordia (Figure 6d,e). These released bacteria did not differentiate, and symbiosomes were not observed (Figure 6d,e).

(c)

(d)

AMSH1 is required downstream of cytokinin signaling

(e)

expressing a hemA::lacZ reporter gene. Dark blue staining was observed at the center of the wild-type nodules, indicating full colonization. In contrast, only weak staining near primordia surfaces was detected in amsh1–1 and amsh1–2, suggesting limited colonization (Figure 6a). Next, we examined the amsh1 primordia and wild-type nodules at 3 and 6 wpi using light and transmission electron microscopy. At 3 wpi, wild-type nodules were fully colonized, with ITs present within the nodules (Figure 6b,

The severe infection defects in the amsh1 mutants may explain the arrested progression of organogenesis observed in response to rhizobia. However, amsh1 mutants may suffer defects in nodule organogenesis independent of their infection deficiencies. To investigate this possibility, we determined the extent of organogenesis induced by exogenous application of cytokinin in the absence of rhizobia and infection (Heckmann et al., 2011). After 6 weeks of treatment with 10
8 M of the cytokinin analog 6–benzylaminopurine (BAP), no apparent growth rate differences were observed between plants growing on medium with and without BAP (Figure S10). BAP treatment induced formation of more than five white nodules per wild-type plant, while only approximately two small primordia per plant were seen in amsh1–1 plants, and very few primordia were induced in amsh1–2 (Figure 7a). The BAP-induced primordia in the amsh1 mutants were much smaller than the white nodulelike structures observed on the wild-type roots (Figure 7b,c). Additionally, we tested whether application of BAP rescued the amsh1–1 and amsh1–2 nodulation phenotype. We

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

AMSH1 promotes infection and organogenesis 725

(a)

(b)

(c)

(a)

(b) (d)

(c)

(d)

(e) (e)

Figure 5. Root hair infection thread and nodule frequencies in wild-type and amsh1 mutants. (a–c) Representative examples of IT categories, i.e. incipient, elongating and long ITs. Scale bars = 50 lm. (d) Number of infection threads of various categories at 8 and 21 dpi. Error bars indicate SEM. Ten plants of each genotype were assayed. (e) Number of nodules and nodule primordia on wild-type and amsh1–1 and amsh1–2 at 3 and 6 wpi. Error bars indicate SEM. Twenty plants of each genotype were assayed.

did not observe a difference in primordium number compared to the plants grown on medium without BAP (Figure S10). The wild-type plants formed approximately three pink nodules, while amsh1 mutants produced only two nodule primordia. Likewise, we found no effect of BAP treatment on the amsh1 infection phenotype, and the cytokinin-treated amsh1–1 and amsh1–2 primordia were not properly colonized, but had a high abundance of branching and bulbous infection threads enclosed in thick cell walllike structures, as previously observed without BAP treatment (Figure 7d–g). In conclusion, amsh1 mutants appear to have organogenesis defects that are independent of infection, suggesting a requirement for AMSH1 downstream of cytokinin signaling.

Figure 6. Nodule infection thread phenotypes of amsh1 mutants. (a) Nodule colonization visualized by lacZ staining (blue) of M. loti expressing a hemA::lacZ reporter gene. Scale bars = 0.5 mm. (b) Micrographs of wild-type nodules and amsh1–1 and amsh1–2 nodule primordia at 3 wpi. Scale bars = 100 lm. Red arrowheads indicate ITs. (c) Transmission electron micrographs of wild-type nodules and amsh1–1 and amsh1–2 nodule primordia at 3 wpi. Scale bars = 2 lm. The green arrow indicates symbiosomes. Infection thread structures are indicated by dashed red lines, and thickened cell-wall structures are indicated by dashed yellow lines. (d) Micrographs of wild-type nodules and amsh1–1 and amsh1–2 nodule primordia at 6 wpi. Scale bars = 100 lm. Red arrowheads indicate ITs. (e) Transmission electron micrographs of wild-type nodules and amsh1–1 and amsh1–2 nodule primordia at 6 wpi. Scale bars = 2 lm. Infection thread structures are indicated by dashed red lines. Green arrows indicate symbiosomes, while blue arrowheads indicate free bacteria. The genotypes are indicated at the bottom of the panels.

DISCUSSION The LORE1 resource facilitated efficient identification and validation of the amsh1 mutants The LORE1 mutant collection represents a substantial reverse genetics resource, providing access to loss-of-function alleles of more than 20 000 unique Lotus genes (Fukai et al., 2012; Urbanski et al., 2012, 2013). Here, we describe the efficient use of the LORE1 resource in a forward genetics approach aimed at identification of novel genetic regulators of nodulation. We performed a LORE1 family screen, whereby multiple mutant individuals displaying the amsh1

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

The Plant Journal (2015) 83, 719–731

doi: 10.1111/tpj.12922

The deubiquitinating enzyme AMSH1 is required for rhizobial infection and nodule organogenesis in Lotus japonicus
ski1, Euan K. James2, Niels Sandal1, Erika Isono3, Jens Stougaard1 and Anna Małolepszy1, Dorian Fabian Urban 1, Stig Uggerhøj Andersen * 1 Centre for Carbohydrate Recognition and Signalling, Department of Molecular Biology and Genetics, Aarhus University, Gustav Wieds Vej 10, DK–8000, Aarhus C, Denmark, 2 The James Hutton Institute, Invergowrie Dundee DD2 5DA, UK, and 3 Department of Plant Systems Biology, Technische Universita€ t Mu€ nchen, D–85354 Freising, Germany Received 22 May 2015; accepted 22 June 2015; published online 28 june 2015. *For correspondence (e-mail [email protected]). Accession numbers: The Amsh1 sequences have been submitted to Genbank under accession number KR270441.

SUMMARY Legume–rhizobium symbiosis contributes large quantities of fixed nitrogen to both agricultural and natural ecosystems. This global impact and the selective interaction between rhizobia and legumes culminating in development of functional root nodules have prompted detailed studies of the underlying mechanisms. We performed a screen for aberrant nodulation phenotypes using the Lotus japonicus LORE1 insertion mutant collection. Here, we describe the identification of amsh1 mutants that only develop small nodule primordia and display stunted shoot growth, and show that the aberrant nodulation phenotype caused by LORE1 insertions in the Amsh1 gene may be separated from the shoot phenotype. In amsh1 mutants, rhizobia initially became entrapped in infection threads with thickened cells walls. Some rhizobia were released into plant cells much later than observed for the wild-type; however, no typical symbiosome structures were formed. Furthermore, cytokinin treatment only very weakly induced nodule organogenesis in amsh1 mutants, suggesting that AMSH1 function is required downstream of cytokinin signaling. Biochemical analysis showed that AMSH1 is an active deubiquitinating enzyme, and that AMSH1 specifically cleaves K63-linked ubiquitin chains. Post-translational ubiquitination and deubiquitination processes involving the AMSH1 deubiquitinating enzyme are thus involved in both infection and organogenesis in Lotus japonicus. Keywords: Lotus japonicus, nodulation mutant, AMSH1, deubiquitination, symbiosis, LORE1.

INTRODUCTION Many legumes are rich sources of protein due to their symbiosis with nitrogen-fixing rhizobia accommodated in root nodules. The symbiotic signaling process is initiated when rhizobia secrete nodulation (Nod) factors upon sensing (iso) flavonoids produced by compatible legumes. The Nod factors consist of N–acetyl-D–glucosamine oligomers linked by b–1,4 bonds, in which the N–acetyl group on the non-reducing end is replaced by a fatty acid (Lerouge et al., 1990; Spaink et al., 1991). Genetic studies of loss-of-function and gain-of-function mutants in Lotus japonicus (Lotus) and Medicago truncatula (Medicago) show that the Lotus Nod factor receptors NFR1 and NFR5 and the corresponding proteins LYK3 and NFP in Medicago (Limpens et al., 2003; Madsen et al., 2003; Radutoiu et al., 2003; Arrighi et al., 2006; Mulder et al., 2006; Smit et al., 2007) are pivotal for perception of rhizobial Nod factors, and NFR1 and NFR5 bind Nod © 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd

factors from compatible bacteria in vitro with high affinity (Broghammer et al., 2012). In the susceptible root zone, Nod factor perception then leads to initiation of a nodulation signaling cascade, which bifurcates into branches promoting epidermal root hair infection and cortical nodule organogenesis (Kouchi et al., 2010; Madsen et al., 2010). Promotion of epidermal infection relies on the activities of the pectate lyase Nodulation pectate lyase (NPL), the ubiquitin ligase CERBERUS, the Nck-associated protein 1 (NAP1), 12IF-specific p53 inducible RNA (PIR1) and Actin-related protein component 1 (ARPC1) proteins required for actin rearrangement, and the transcription factors Nodule inception protein (NIN), Nodulation signaling pathway 1 (NSP1) and Nodulation signaling pathway 2 (NSP2) (Schauser et al., 1999; Kalo et al., 2005; Smit et al., 2005; Heckmann et al., 2006; Murakami et al., 2006; Marsh et al., 2007; Yano et al., 2008, 2009; Yokota et al., 2009; Hossain et al., 2012; Xie et al., 2012). 719

AMSH1 promotes infection and organogenesis 727 comparable to the wild-type, and produce lateral roots (Figure 3). Despite this overall normal appearance of amsh1 roots, they fail to produce functional nodules. Instead, amsh1 displays impaired IT progression and bacterial release from the ITs, and amsh1 nodule development is arrested at the primordium stage. AMSH1 may target nodulation proteins The most likely explanation for these observations is that one or more critical components of the nodulation signaling machinery are highly sensitive to K63-linked ubiquitination, and require AMSH1 DUB activity to retain functionality. The simplest hypothesis is that AMSH1 promotes infection and organogenesis by acting on a single nodulation component, leaving the transcriptional regulators NIN, NSP1 and NSP2 as candidate AMSH1 targets, because they are required for both processes (Madsen et al., 2010; Heckmann et al., 2011). The remaining nodulation proteins are less likely candidates under this hypothesis, because they either have more specific infection- or organogenesis-related functions, or because their corresponding loss-of-function mutants, unlike amsh1, respond to cytokinin treatment by formation of larger nodule-like structures. The infection mutants nap, pir, arpC1, cyclops and cerberus thus display abnormal IT development, but nodule organogenesis proceeds further than in amsh1. Conversely, loss of organogenesis-specific proteins, such as the cytokinin receptor LHK1 (Tirichine et al., 2006b, 2007) and the DNA topoisomerase VI components VAG1 and SUNER1 (Suzaki et al., 2014; Yoon et al., 2014), does not hinder epidermal infection, but results in reduced or absent organogenesis. The nfr1, nfr5, symrk, nup133, nup85, castor, pollux, ccamk, pir and cyclops mutants all respond to cytokinin treatment with formation of nodule-like structures (Heckmann et al., 2011), making it unlikely that AMSH1 targeting of one of the corresponding proteins explains the requirement for AMSH1 downstream of cytokinin signaling. An alternative hypothesis to targeting a single nodulation component required for both infection and organogenesis is that AMSH1 targets multiple components within both signaling pathways, for instance LHK1 and one or more of the proteins required for infection. With the availability of multiple amsh1 mutant alleles, these hypotheses may now be tested using biochemical approaches to compare the ubiquitination status and protein abundance of known nodulation signaling components in amsh1 and wild-type backgrounds. The AMSH1 characterization presented here thus opens new avenues of investigation with the potential for greatly advancing the understanding of the effect of ubiquitination on nodulation signaling in legumes.

EXPERIMENTAL PROCEDURES Bacterial strains Mesorhizobium loti strain NZP2235 was used for forward screening under greenhouse conditions, but for most other studies, M. loti strain MAFF303099 expressing DsRed was used (Maekawa et al., 2009). The M. loti strain NZP2235 expressing hemA::lacZ reporter gene was used for lacZ staining (Schauser et al., 1998; Wopereis et al., 2000). Cultures with an OD600 of 0.06 were used for all studies involving rhizobia. Agrobacterium rhizogenes strain AR1193 was used for hairy root transformation (Stougaard et al., 1987).

Plant materials amsh1 mutants were obtained from the LORE1 collection (Fukai et al., 2012; Urbanski et al., 2012). The isolation of nfr1–1 and nfr5–2 has been described by Schauser et al. (1998). Lotus japonicus B–129 Gifu is the wild-type for all these mutants (Handberg and Stougaard, 1992). Seeds were scarified by treatment with 98% sulfuric acid for 15 min, or using sand paper, and then sterilized with 0.5% sodium hypochlorite for 20 min. They were rinsed five times with sterile water, and incubated on a shaker for 3 h at room temperature. Seeds were germinated on moist, sterile paper on vertical plates in growth chambers under the following conditions: 16/8 h light/darkness at 21°C/18°C for 3 days. For nodulation tests, they were moved to plates containing quarter-strength B&D medium (Broughton and Dilworth, 1971) and 1.4% Noble agar (Difco, http:// www.bd.com) without nitrate, or to Magenta containers (Sigma, http://www.sigmaaldrich.com) filled with lightweight expanded clay aggregate (LECA, 2-4 mm diameter, Saint-Gobain Weber A/S, http://www.weber.dk) and vermiculite size M (Damolin A/S http:// www.damolin.dk) in a 4:1 mixture, and 80 ml of quarter-strength B&D medium without nitrate. On plates, roots were shielded from light using a metal spacer. For cytokinin experiments, quarterstrength B&D medium was supplemented with BAP to a final concentration of 10
8 M. For complementation, 3-day-old seedlings were moved to half-strength Gamborg’s B5 basal salt mixture (Duchefa Biochemie, https://www.duchefa-biochemie.com) supplemented with Gamborg’s vitamin solution (Sigma). For each experiment, a segregating population of amsh1 plants was used. Preparation of 200 seeds usually yielded approximately 20 amsh1 plants. For root growth assays, 20 Gifu plants, 17 amsh1–1 plants and 15 amsh1–2 plants were assayed. For nodule counts, 20 plants from each genotype were used. For infection thread counts, ten plants from each genotype were used at 8 and 21 dpi. For Western blots, the protein extract from one seedling was used, and blotting was repeated three times. For hairy root complementation, 1000 seeds from a segregating population were used: 500 seeds were transformed with empty vector control and 500 with 35S::Amsh1 cDNA. After 4 weeks, the small plants were moved to Magenta containers and inoculated. Six weeks later, plants were scored, photographs were taken and plants were genotyped.

Arbuscular mycorrhiza One-week-old plants from a segregating population of amsh1 and wild-type Gifu seedlings were transferred to Rhizophagus irregularis-colonized chive nurse pots as previously described (Chabot et al., 1992; Kosuta et al., 2005; Kruger et al., 2012). Plants were co-cultivated for 5–6 weeks at 25°C. Each week, they were fertilized with chive nutrient solution supplemented with potassium nitrate to a final concentration of 5.5 mM nitrate. Plants roots were

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

728 Anna Małolepszy et al. scored for arbuscular mycorrhizal colonization by ink staining as previously described (Vierheilig et al., 1998).

Genotyping DNA was extracted according to the standard CTAB method (Rogers and Bendich, 1985). The genotyping PCR reactions were performed using primers designed using FSTpoolit software (http://carb.au.dk/resources/) as previously described (Urbanski et al., 2012), except that the touch-down step was omitted and a three-step 30-cycle reaction was performed with a 62°C annealing temperature.

Cloning procedure The Lotus Amsh1 coding sequence was cloned from a Lotus cDNA library obtained from roots using F1 and R1 primers (Table S2). The promoter region was amplified together with the 5’UTR from genomic DNA of Gifu using F2 and R2 primers. The promoter region and cDNA were combined using F3 and R3 primers. Phusion polymerase (Thermo Scientific, https://www.lifetechnologies.com/dk/en/home/brands/thermo-scientific.html) was used to amplify all of the fragments. They were cloned into pENTR/d–TOPO vector (Invitrogen, http://www.lifetechnologies.com/dk/en/home/brands/invitrogen.html) between the attL1 and attL2 sites. An LR reaction was performed according to the manufacturer’s instructions in order to transfer the constructs from pENTR/d–TOPO to a pIV10 expression vector containing a Gateway cassette (Stougaard et al., 1987). Ligation-independent cloning was performed according to the Novagen, http://www.novagen.com/ to clone the Amsh1 cDNA into pET41 (Novagen) for recombinant expression in E. coli. The F4 and R4 primers were used to amplify the Amsh1 cDNA (Table S2).

Quantitative real-time –PCR analysis Total RNA was isolated from various tissues 3 weeks after inoculation with M. loti MAFF303099 or mock-inoculated using the modified lithium chloride/TRIzol LS method as described by Holt et al. (2015). cDNA was synthesized using an oligo(dT) primer and the MuLV reverse transcriptase (Thermo Scientific) as described by Holt et al. (2015). The F5 and R5 primers were used to amplify Amsh1 transcripts. LjUBC, encoding a ubiquitin-conjugating enzyme, was used as a reference gene. The primers used for LjUBC were F6 and R6 (Table S2). Real-time PCR was performed using a LightCycler 480 II (Roche, https://lifescience.roche.com/) with LightCycler 480 SYBR Green I Master mix (Roche).

Protein extraction, DUB assays, and Western blots The GST–AMSH1 fusion protein was expressed from a pET41 vector in E. coli Rosetta strain (DE3) (Merck Chemicals, http:// www.merckmillipore.com) at 18°C, and purified using glutathione Sepharose 4B beads (GE Healthcare, http://www3.gehealthcare.com). Protein was eluted from the beads by incubation for 30 min at 4°C with 100 mM glutathione. DUB assays were performed as described by Isono (2010). LjAMSH1 enzyme (8 pmol) in DUB buffer were incubated with 125 ng ll
1 of K48-, K63-linked or linear ubiquitin chains (two to seven ubiquitins, Enzo Life Sciences, http://www.enzolifesciences.com) for 120 min at 30°C. The reaction was then stopped by adding 2.5 ll of 49 LDS (Lithium dodecyl sulfate) sample buffer (Invitrogen). The samples were run on a NuPAGE Novex Bis/Tris 4-

12% gel (Invitrogen) in MES buffer according to the manufacturer’s instructions. The antibodies used were anti-GST (1:2000, raised in goat, GE Healthcare, http://www3.gehealthcare.com), anti-CDC2 kinase (1:5000, raised in rabbit, Santa Cruz Technology, http://www.scbt.com), anti-Ub P4D1 (1:2500, raised in mouse, Cell Signaling Technology, http://www.cellsignal.com), anti-mouse horseradish peroxidase-conjugated antibody (1:2000, Pierce, http://www.pierce-antibodies.com), anti-goat alkaline phosphatase-conjugated antibody (1:1000, Sigma), and anti-rabbit alkaline phosphatase-conjugated antibody (1:4000, Sigma). SuperSignal West Femto maximum sensitivity substrate (Thermo Scientific) was used for horseradish peroxidase reaction. The 5-bromo-4-chloro-30 -indolyphosphate and nitro-blue tetrazolium chloride for alkaline phosphatase detection were purchased from AppliChem (https://www.applichem.com/).

Microscopic observation, staining and image processing Characterization of nodulation phenotypes was performed using M. loti MAFF303099 expressing DsRed. A Zeiss (http://www.zeiss.com) 780LSM confocal microscope was used for counting of ITs. ITs were counted at 8 and 21 dpi. To visualize nodule colonization, plants were inoculated with M. loti NZP2235 expressing hemA:: lacZ, and lacZ staining was performed as previously described (Wopereis et al., 2000). The root lengths were measured using ImageJ (http://imagej.nih.gov/ij/). Nodules/primordia at 6 wpi were sectioned using a Leica (http://www.leica.com) vibratome. Transmission electron microscopy was performed on ultra-thin sections (80 nm thick) of fixed and resin-embedded nodules and nodule primordia as described by Madsen et al. (2010), and samples viewed and photographed using a JEOL (http://www.jeol.co.jp/en/) JEM1400 TEM.

ACKNOWLEDGMENTS This work was funded by Deutsche Forschungsgemeinschaft grant number IS 221/2–2 (SPP1365) to E.I., Danish National Research Foundation grant number DNRF79 to J.S., European Research Council advanced grant number 268523 to J.S., and EMBO shortterm fellowship number ASTF-408-2012 to A.M. The authors would like to thank Dennis Berg Holt and Katharina Markmann (Centre for Carbohydrate Recognition and Signalling, Department of Molecular Biology and Genetics, Aarhus University, Gustav Wieds Vej 10, DK–8000, Aarhus C, Denmark) for supplying the cDNA used for Amsh1 transcription analysis.

AUTHOR CONTRIBUTIONS A.M., D.F.U. and N.S. performed the forward screening. D.F.U. and A.M. performed the cloning and the sequencespecific amplification polymorphism analysis. A.M. characterized the mutant. E.K.J. performed the light and transmission electron microscopy. E.I. supervised the biochemical assays. S.U.A and J.S. supervised the work. A.M. and S.U.A wrote the manuscript with input from J.S. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article. Figure S1. Alignment of Lotus and Arabidopsis AMSH homologs. Figure S2. amsh1–1 and amsh1–2 are allelic. Figure S3. Phenotypes of the amsh1–3 to amsh1–8 alleles.

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

AMSH1 promotes infection and organogenesis 729 Figure S4. Alignment of human AMSH (HsAMSH), Arabidopsis AMSH1 (AtAMSH1) and Lotus AMSH1 (LjAMSH) protein sequences. Figure S5. Relative expression of Amsh1 in roots, nodules and leaves of inoculated and uninoculated wild-type plants. Figure S6. Comparison of shoot and root growth in the wild-type and amsh1 mutants. Figure S7. amsh1 mutants are able to establish arbuscular mycorrhiza with Rhizophagus irregularis. Figure S8. Root-specific complementation restores normal infection in amsh1–1. Figure S9. amsh1–1 and amsh1–2 mutants respond to bacteria but fail to form long infection threads Figure S10. Application of exogenous cytokinin does not reverse the amsh1 shoot and nodulation phenotype. Table S1. Complementation of the amsh1–1 mutant. Table S2. Primers used in this study.

REFERENCES Amerik, A.Y. and Hochstrasser, M. (2004) Mechanism and function of deubiquitinating enzymes. Biochim. Biophys. Acta, 1695, 189–207. Amerik, A.Y., Swaminathan, S., Krantz, B.A., Wilkinson, K.D. and Hochstrasser, M. (1997) In vivo disassembly of free polyubiquitin chains by yeast Ubp14 modulates rates of protein degradation by the proteasome. EMBO J. 16, 4826–4838. Arrighi, J.F., Barre, A., Ben Amor, B. et al. (2006) The Medicago truncatula lysine motif receptor-like kinase gene family includes NFP and new nodule-expressed genes. Plant Physiol. 142, 265–279. Broghammer, A., Krusell, L., Blaise, M. et al. (2012) Legume receptors perceive the rhizobial lipochitin oligosaccharide signal molecules by direct binding. Proc. Natl Acad. Sci. USA, 109, 13859–13864. Broughton, W.J. and Dilworth, M.J. (1971) Control of leghaemoglobin synthesis in snake beans. Biochem. J. 125, 1075–1080. Chabot, S., Becard, G. and Piche, Y. (1992) Life cycle of Glomus intraradix in root organ culture. Mycologia, 84, 315–321. Den Herder, G., De Keyser, A., De Rycke, R. et al. (2008) Seven in Absentia proteins affect plant growth and nodulation in Medicago truncatula. Plant Physiol. 148, 369–382. Den Herder, G., Yoshida, S., Antolin-Llovera, M., Ried, M.K. and Parniske, M. (2012) Lotus japonicus E3 ligase SEVEN IN ABSENTIA4 destabilizes the symbiosis receptor-like kinase SYMRK and negatively regulates rhizobial infection. Plant Cell, 24, 1691–1707. Endre, G., Kereszt, A., Kevei, Z., Mihacea, S., Kalo, P. and Kiss, G.B. (2002) A receptor kinase gene regulating symbiotic nodule development. Nature, 417, 962–966. Fukai, E., Soyano, T., Umehara, Y., Nakayama, S., Hirakawa, H., Tabata, S., Sato, S. and Hayashi, M. (2012) Establishment of a Lotus japonicus gene tagging population using the exon-targeting endogenous retrotransposon LORE1. Plant J. 69, 720–730. Handberg, K. and Stougaard, J. (1992) Lotus japonicus, an autogamous, diploid legume species for classical and molecular genetics. Plant J. 2, 487–496. Heckmann, A.B., Lombardo, F., Miwa, H., Perry, J.A., Bunnewell, S., Parniske, M., Wang, T.L. and Downie, J.A. (2006) Lotus japonicus nodulation requires two GRAS domain regulators, one of which is functionally conserved in a non-legume. Plant Physiol. 142, 1739–1750. Heckmann, A.B., Sandal, N., Bek, A.S., Madsen, L.H., Jurkiewicz, A., Nielsen, M.W., Tirichine, L. and Stougaard, J. (2011) Cytokinin induction of root nodule primordia in Lotus japonicus is regulated by a mechanism operating in the root cortex. Mol. Plant-Microbe Interact. 24, 1385–1395. Herve
, C., Lefebvre, B. and Cullimore, J. (2011) How many E3 ubiquitin ligases are involved in the regulation of nodulation? Plant Signal. Behav. 6, 660–664. Holt, D.B., Gupta, V., Meyer, D., Abel, N.B., Andersen, S.U., Stougaard, J. and Markmann, K. (2015) micro RNA 172 (miR172) signals epidermal infection

and is expressed in cells primed for bacterial invasion in Lotus japonicus roots and nodules. New Phytol. doi:10.1111/nph.13445 [Epub ahead of print]. Hossain, M.S., Liao, J., James, E.K., Sato, S., Tabata, S., Jurkiewicz, A., Madsen, L.H., Stougaard, J., Ross, L. and Szczyglowski, K. (2012) Lotus japonicus ARPC1 is required for rhizobial infection. Plant Physiol. 160, 917–928. Imaizumi-Anraku, H., Takeda, N., Charpentier, M. et al. (2005) Plastid proteins crucial for symbiotic fungal and bacterial entry into plant roots. Nature, 433, 527–531. Isono, E. and Nagel, M.K. (2014) Deubiquitylating enzymes and their emerging role in plant biology. Front. Plant Sci. 5, 56. Isono, E., Katsiarimpa, A., Muller, I.K., Anzenberger, F., Stierhof, Y.D., Geldner, N., Chory, J. and Schwechheimer, C. (2010) The deubiquitinating enzyme AMSH3 is required for intracellular trafficking and vacuole biogenesis in Arabidopsis thaliana. Plant Cell, 22, 1826–1837. Kalo, P., Gleason, C., Edwards, A. et al. (2005) Nodulation signaling in legumes requires NSP2, a member of the GRAS family of transcriptional regulators. Science, 308, 1786–1789. Kanamori, N., Madsen, L.H., Radutoiu, S. et al. (2006) A nucleoporin is required for induction of Ca2+ spiking in legume nodule development and essential for rhizobial and fungal symbiosis. Proc. Natl Acad. Sci. USA, 103, 359–364. Katsiarimpa, A., Anzenberger, F., Schlager, N., Neubert, S., Hauser, M.T., Schwechheimer, C. and Isono, E. (2011) The Arabidopsis deubiquitinating enzyme AMSH3 interacts with ESCRT–III subunits and regulates their localization. Plant Cell, 23, 3026–3040. Katsiarimpa, A., Kalinowska, K., Anzenberger, F., Weis, C., Ostertag, M., Tsutsumi, C., Schwechheimer, C., Brunner, F., Huckelhoven, R. and Isono, E. (2013) The deubiquitinating enzyme AMSH1 and the ESCRT–III subunit VPS2.1 are required for autophagic degradation in Arabidopsis. Plant Cell, 25, 2236–2252. Katsiarimpa, A., Munoz, A., Kalinowska, K., Uemura, T., Rojo, E. and Isono, E. (2014) The ESCRT–III-interacting deubiquitinating enzyme AMSH3 is essential for degradation of ubiquitinated membrane proteins in Arabidopsis thaliana. Plant Cell Physiol. 55, 727–736. Kawaguchi, M., Imaizumi-Anraku, H., Koiwa, H., Niwa, S., Ikuta, A., Syono, K. and Akao, S. (2002) Root, root hair, and symbiotic mutants of the model legume Lotus japonicus. Mol. Plant-Microbe Interact. 15, 17–26. Kimura, Y. and Tanaka, K. (2010) Regulatory mechanisms involved in the control of ubiquitin homeostasis. J. Biochem. 147, 793–798. Komander, D. and Rape, M. (2012) The ubiquitin code. Annu. Rev. Biochem. 81, 203–229. Kosuta, S., Winzer, T. and Parniske, M. (2005) Arbuscular mycorrhiza. In Lotus Japonicus Handbook (Marquez, A.J., ed.). Dordrecht, The Netherlands: Springer, pp. 87–95. Kouchi, H., Imaizumi-Anraku, H., Hayashi, M., Hakoyama, T., Nakagawa, T., Umehara, Y., Suganuma, N. and Kawaguchi, M. (2010) How many peas in a pod? Legume genes responsible for mutualistic symbioses underground. Plant Cell Physiol. 51, 1381–1397. Kruger, M., Kruger, C., Walker, C., Stockinger, H. and Schussler, A. (2012) Phylogenetic reference data for systematics and phylotaxonomy of arbuscular mycorrhizal fungi from phylum to species level. New Phytol. 193, 970–984. Lerouge, P., Roche, P., Faucher, C., Maillet, F., Truchet, G., Prome, J.C. and Denarie, J. (1990) Symbiotic host-specificity of Rhizobium meliloti is determined by a sulphated and acylated glucosamine oligosaccharide signal. Nature, 344, 781–784. Levy, J., Bres, C., Geurts, R. et al. (2004) A putative Ca2+ and calmodulin-dependent protein kinase required for bacterial and fungal symbioses. Science, 303, 1361–1364. Liao, J., Singh, S., Hossain, M.S. et al. (2012) Negative regulation of CCaMK is essential for symbiotic infection. Plant J. 72, 572–582. Limpens, E., Franken, C., Smit, P., Willemse, J., Bisseling, T. and Geurts, R. (2003) LysM domain receptor kinases regulating rhizobial Nod factor-induced infection. Science, 302, 630–633. ~ ez, A., Soto, T., Vicente-Soler, J., Gacto, M. and Cansado, Madrid, M., N
un J. (2007) Stress-activated protein kinase-mediated down-regulation of the cell integrity pathway mitogen-activated protein kinase Pmk1p by protein phosphatases. Mol. Biol. Cell, 18, 4405–4419.

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

730 Anna Małolepszy et al. Madsen, E.B., Madsen, L.H., Radutoiu, S. et al. (2003) A receptor kinase gene of the LysM type is involved in legume perception of rhizobial signals. Nature, 425, 637–640. Madsen, L.H., Tirichine, L., Jurkiewicz, A., Sullivan, J.T., Heckmann, A.B., Bek, A.S., Ronson, C.W., James, E.K. and Stougaard, J. (2010) The molecular network governing nodule organogenesis and infection in the model legume Lotus japonicus. Nat. Commun. 1, 10. Maekawa, T., Maekawa-Yoshikawa, M., Takeda, N., Imaizumi-Anraku, H., Murooka, Y. and Hayashi, M. (2009) Gibberellin controls the nodulation signaling pathway in Lotus japonicus. Plant J. 58, 183–194. Marsh, J.F., Rakocevic, A., Mitra, R.M., Brocard, L., Sun, J., Eschstruth, A., Long, S.R., Schultze, M., Ratet, P. and Oldroyd, G.E. (2007) Medicago truncatula NIN is essential for rhizobial-independent nodule organogenesis induced by autoactive calcium/calmodulin-dependent protein kinase. Plant Physiol. 144, 324–335. Maytal-Kivity, V., Reis, N., Hofmann, K. and Glickman, M.H. (2002) MPN+, a putative catalytic motif found in a subset of MPN domain proteins from eukaryotes and prokaryotes, is critical for Rpn11 function. BMC Biochem. 3, 28. Mbengue, M., Camut, S., de Carvalho-Niebel, F. et al. (2010) The Medicago truncatula E3 ubiquitin ligase PUB1 interacts with the LYK3 symbiotic receptor and negatively regulates infection and nodulation. Plant Cell, 22, 3474–3488. McCullough, J., Clague, M.J. and Urbe, S. (2004) AMSH is an endosome-associated ubiquitin isopeptidase. J. Cell Biol. 166, 487–492. Mitra, R.M., Gleason, C.A., Edwards, A., Hadfield, J., Downie, J.A., Oldroyd, G.E. and Long, S.R. (2004) A Ca2+/calmodulin-dependent protein kinase required for symbiotic nodule development: gene identification by transcript-based cloning. Proc. Natl Acad. Sci. USA, 101, 4701–4705. Mukhopadhyay, D. and Riezman, H. (2007) Proteasome-independent functions of ubiquitin in endocytosis and signaling. Science, 315, 201– 205. Mulder, L., Lefebvre, B., Cullimore, J. and Imberty, A. (2006) LysM domains of Medicago truncatula NFP protein involved in Nod factor perception. Glycosylation state, molecular modeling and docking of chitooligosaccharides and Nod factors. Glycobiology, 16, 801– 809. Murakami, Y., Miwa, H., Imaizumi-Anraku, H., Kouchi, H., Downie, J.A., Kawaguchi, M. and Kawasaki, S. (2006) Positional cloning identifies Lotus japonicus NSP2, a putative transcription factor of the GRAS family, required for NIN and ENOD40 gene expression in nodule initiation. DNA Res. 13, 255–265. Murray, J.D., Karas, B.J., Sato, S., Tabata, S., Amyot, L. and Szczyglowski, K. (2007) A cytokinin perception mutant colonized by Rhizobium in the absence of nodule organogenesis. Science, 315, 101–104. Murray, J.D., Muni, R.R., Torres-Jerez, I. et al. (2011) Vapyrin, a gene essential for intracellular progression of arbuscular mycorrhizal symbiosis, is also essential for infection by rhizobia in the nodule symbiosis of Medicago truncatula. Plant J. 65, 244–252. Nijman, S.M., Luna-Vargas, M.P., Velds, A., Brummelkamp, T.R., Dirac, A.M., Sixma, T.K. and Bernards, R. (2005) A genomic and functional inventory of deubiquitinating enzymes. Cell, 123, 773–786. Radutoiu, S., Madsen, L.H., Madsen, E.B. et al. (2003) Plant recognition of symbiotic bacteria requires two LysM receptor-like kinases. Nature, 425, 585–592. Riabowol, K., Draetta, G., Brizuela, L., Vandre, D. and Beach, D. (1989) The cdc2 kinase is a nuclear protein that is essential for mitosis in mammalian cells. Cell, 57, 393–401. Rogers, S.O. and Bendich, A.J. (1985) Extraction of DNA from milligram amounts of fresh, herbarium and mummified plant tissues. Plant Mol. Biol. 5, 69–76. Saito, K., Yoshikawa, M., Yano, K. et al. (2007) NUCLEOPORIN85 is required for calcium spiking, fungal and bacterial symbioses, and seed production in Lotus japonicus. Plant Cell, 19, 610–624. Sato, Y., Yoshikawa, A., Yamagata, A., Mimura, H., Yamashita, M., Ookata, K., Nureki, O., Iwai, K., Komada, M. and Fukai, S. (2008) Structural basis for specific cleavage of Lys 63-linked polyubiquitin chains. Nature, 455, 358–362. Schauser, L., Handberg, K., Sandal, N., Stiller, J., Thykjaer, T., Pajuelo, E., Nielsen, A. and Stougaard, J. (1998) Symbiotic mutants deficient in nod-

ule establishment identified after T-DNA transformation of Lotus japonicus. Mol. Gen. Genet. 259, 414–423. Schauser, L., Roussis, A., Stiller, J. and Stougaard, J. (1999) A plant regulator controlling development of symbiotic root nodules. Nature, 402, 191– 195. Singh, S., Katzer, K., Lambert, J., Cerri, M. and Parniske, M. (2014) CYCLOPS, a DNA-binding transcriptional activator, orchestrates symbiotic root nodule development. Cell Host Microbe. 15, 139–152. Smit, P., Raedts, J., Portyanko, V., Debelle, F., Gough, C., Bisseling, T. and Geurts, R. (2005) NSP1 of the GRAS protein family is essential for rhizobial Nod factor-induced transcription. Science, 308, 1789– 1791. Smit, P., Limpens, E., Geurts, R., Fedorova, E., Dolgikh, E., Gough, C. and Bisseling, T. (2007) Medicago LYK3, an entry receptor in rhizobial nodulation factor signaling. Plant Physiol. 145, 183–191. Spaink, H.P., Sheeley, D.M., van Brussel, A.A.N., Glushka, J., York, W.S., Tak, T., Geiger, O., Kennedy, E.P., Reinhold, V.N. and Lugtenberg, B.J.J. (1991) A novel highly unsaturated fatty acid moiety of lipo-oligosaccharide signals determines host specificity of Rhizobium. Nature, 354, 125– 130. Stougaard, J., Abildsten, D. and Marcker, K.A. (1987) The Agrobacterium rhizogenes pRi TL-DNA segment as a gene vector system for transformation of plants. Mol. Gen. Genet. 207, 251–255. Stracke, S., Kistner, C., Yoshida, S. et al. (2002) A plant receptor-like kinase required for both bacterial and fungal symbiosis. Nature, 417, 959–962. Suzaki, T., Ito, M., Yoro, E., Sato, S., Hirakawa, H., Takeda, N. and Kawaguchi, M. (2014) Endoreduplication-mediated initiation of symbiotic organ development in Lotus japonicus. Development, 141, 2441–2445. Szczyglowski, K., Shaw, R.S., Wopereis, J., Copeland, S., Hamburger, D., Kasiborski, B., Dazzo, F.B. and de Bruijn, F.J. (1998) Nodule organogenesis and symbiotic mutants of the model legume Lotus japonicus. Mol. Plant-Microbe Interact. 11, 684–697. Tirichine, L., Imaizumi-Anraku, H., Yoshida, S. et al. (2006a) Deregulation of a Ca2+/calmodulin-dependent kinase leads to spontaneous nodule development. Nature, 441, 1153–1156. Tirichine, L., James, E.K., Sandal, N. and Stougaard, J. (2006b) Spontaneous root-nodule formation in the model legume Lotus japonicus: a novel class of mutants nodulates in the absence of rhizobia. Mol. PlantMicrobe Interact. 19, 373–382. Tirichine, L., Sandal, N., Madsen, L.H., Radutoiu, S., Albrektsen, A.S., Sato, S., Asamizu, E., Tabata, S. and Stougaard, J. (2007) A gain-of-function mutation in a cytokinin receptor triggers spontaneous root nodule organogenesis. Science, 315, 104–107. Tran, H.J.T.T., Allen, M.D., Lowe, J. and Bycroft, M. (2003) Structure of the Jab1/MPN domain and its implications for proteasome function. Biochemistry, 42, 11460–11465. Urbanski, D.F., Malolepszy, A., Stougaard, J. and Andersen, S.U. (2012) Genome-wide LORE1 retrotransposon mutagenesis and high-throughput insertion detection in Lotus japonicus. Plant J. 69, 731–741. Urbanski, D.F., Malolepszy, A., Stougaard, J. and Andersen, S.U. (2013) High-throughput and targeted genotyping of Lotus japonicus LORE1 insertion mutants. In Legume Genomics: Methods and Protocols. Methods in Molecular Biology Volume 1069 (Rose, R.J., ed.). Totowa, NJ: Humana Press, pp. 119–146. Vierheilig, H., Coughlan, A.P., Wyss, U. and Piche, Y. (1998) Ink and vinegar, a simple staining technique for arbuscular-mycorrhizal fungi. Appl. Environ. Microbiol. 64, 5004–5007. Wilkinson, K.D., Tashayev, V.L., O’Connor, L.B., Larsen, C.N., Kasperek, E. and Pickart, C.M. (1995) Metabolism of the polyubiquitin degradation signal: structure, mechanism, and role of isopeptidase T. Biochemistry, 34, 14535–14546. Woelk, T., Sigismund, S., Penengo, L. and Polo, S. (2007) The ubiquitination code: a signalling problem. Cell Div. 2, 11. Wopereis, J., Pajuelo, E., Dazzo, F.B., Jiang, Q., Gresshoff, P.M., de Bruijn, F.J., Stougaard, J. and Szczyglowski, K. (2000) Short root mutant of Lotus japonicus with a dramatically altered symbiotic phenotype. Plant J. 23, 97–114. Xie, F., Murray, J.D., Kim, J., Heckmann, A.B., Edwards, A., Oldroyd, G.E. and Downie, J.A. (2012) Legume pectate lyase required for root infection by rhizobia. Proc. Natl Acad. Sci. USA, 109, 633–638.

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731

AMSH1 promotes infection and organogenesis 731 Yano, K., Yoshida, S., Muller, J. et al. (2008) CYCLOPS, a mediator of symbiotic intracellular accommodation. Proc. Natl Acad. Sci. USA, 105, 20540– 20545. Yano, K., Shibata, S., Chen, W.L. et al. (2009) CERBERUS, a novel U-box protein containing WD-40 repeats, is required for formation of the infection thread and nodule development in the legume–rhizobium symbiosis. Plant J. 60, 168–180. Ye, Y. and Rape, M. (2009) Building ubiquitin chains: E2 enzymes at work. Nat. Rev. Mol. Cell Biol. 10, 755–764.

Yokota, K., Fukai, E., Madsen, L.H. et al. (2009) Rearrangement of actin cytoskeleton mediates invasion of Lotus japonicus roots by Mesorhizobium loti. Plant Cell, 21, 267–284. Yoon, H.J., Hossain, M.S., Held, M. et al. (2014) Lotus japonicus SUNERGOS1 encodes a predicted subunit A of a DNA topoisomerase VI that is required for nodule differentiation and accommodation of rhizobial infection. Plant J. 78, 811–821. Yuan, S., Zhu, H., Gou, H. et al. (2012) A ubiquitin ligase of symbiosis receptor kinase involved in nodule organogenesis. Plant Physiol. 160, 106–117.

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 83, 719–731